Framing Charitable Donations as Exceptional ... - Faculty & Research

Journal of Experimental Psychology: Applied Framing Charitable Donations as Exceptional Expenses Increases Giving Abigail B. Sussman, Eesha Sharma, and Adam L. Alter Online First Publication, April 20, 2015. http://dx.doi.org/10.1037/xap0000047

CITATION Sussman, A. B., Sharma, E., & Alter, A. L. (2015, April 20). Framing Charitable Donations as Exceptional Expenses Increases Giving. Journal of Experimental Psychology: Applied. Advance online publication. http://dx.doi.org/10.1037/xap0000047

Journal of Experimental Psychology: Applied 2015, Vol. 21, No. 2, 000

© 2015 American Psychological Association 1076-898X/15/$12.00 http://dx.doi.org/10.1037/xap0000047

Framing Charitable Donations as Exceptional Expenses Increases Giving Abigail B. Sussman

Eesha Sharma

University of Chicago

Dartmouth College

Adam L. Alter

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

New York University Many articles have examined the psychological drivers of charitable giving, but little is known about how people mentally budget for charitable gifts. The present research aims to address this gap by investigating how perceptions of donations as exceptional (uncommon and infrequent) rather than ordinary (common and frequent) expenses might affect budgeting for and giving to charity. We provide the first demonstration that exceptional framing of an identical item can directly influence mental budgeting processes, and yield societal benefits. In 5 lab and field experiments, exceptional framing increased charitable behavior, and diminished the extent to which people considered the effect of the donation on their budgets. The current work extends our understanding of mental accounting and budgeting for charitable gifts, and demonstrates practical techniques that enable fundraisers to enhance the perceived exceptionality of donations. Keywords: charitable giving, exceptional expenses, choice bracketing, mental accounting

charitable giving, and how charitable appeals might influence the budgeting process. However, recent research has revealed that people do in fact have mental budgets for philanthropy and that these budgets are malleable (LaBarge & Stinson, 2014). A better understanding of how donations interact with mental budgets might prove particularly useful, since most people have limited funds and must consider their budgets when deciding how much to spend on charitable gifts (Margolis, 1984). The present research aims to address this gap in the literature by understanding how one aspect of mental accounting—whether donations are perceived as exceptional (uncommon or infrequent) or ordinary (common or frequent) expenses—affects budgeting for and giving to charitable causes. This article provides three central contributions. First, it demonstrates that framing an item as exceptional makes people less likely to consider their budget and more likely to donate a greater amount. Second, it shows that minimal variation in the described frequency of an identical charitable donation (e.g., annual vs. once a year) can be sufficient to alter perceived exceptionality and increase willingness to give. Third, it provides charity campaign managers with an inexpensive and easy-to-implement technique that might encourage people to donate more generously to existing charity campaigns.

Annual charitable giving in the United States remains 8% lower than before the Great Recession in 2007 (Giving USA, 2013). The recovery in donations has been more sluggish than expected, in part because recent economic events have forced people to consider their budgets more carefully when deciding whether and how much to give (Linn, 2013). While many articles have examined the psychological drivers of charitable giving (for overviews, see Bendapudi, Singh, & Bendapudi, 1996; Oppenheimer & Olivola, 2011), most sought to understand these decisions in terms of the benefits people derive from giving (e.g., Andreoni, 1989, 1990; Dunn, Aknin, & Norton, 2008; Glazer & Konrad, 1996; Griskevicius et al., 2007; Strahilevitz, 2011), or the emotions they experience when asked to give (e.g., Batson, 1990; Dickert, Sagara, & Slovic, 2011; Small, 2011). This work has expanded our knowledge of how people’s values, motives, and affective reactions influence whether and how much they donate to charitable appeals. In contrast, relatively little is known about how people budget for

Abigail B. Sussman, Booth School of Business, University of Chicago; Eesha Sharma, Tuck School of Business, Dartmouth College; Adam L. Alter, Leonard N. Stern School of Business, New York University. The first two authors contributed equally. This work was supported by the Beatrice Foods Co. Faculty Research Fund at the University of Chicago Booth School of Business. We thank conference participants at the Society for Judgment and Decision Making (SJDM) and the Behavioral Decision Research in Management (BDRM) conferences for useful suggestions, and Belinda Eg for helpful research assistance. Correspondence concerning this article should be addressed to Abigail B. Sussman, University of Chicago Booth School of Business, 5807 S. Woodlawn Avenue, Chicago, IL 60637, or Eesha Sharma, Dartmouth College Tuck School of Business, 100 Tuck Hall, Hanover, NH 03755. E-mail: [email protected] or [email protected]

Mental Accounting for Charitable Donations Mental accounting research suggests that budgeting is a twostage process involving both booking and posting (Heath, 1995; Heath & Soll, 1996). Booking an expense requires that people take notice of and record the expense in their accounting system; posting an expense requires that they assign costs to specific accounts. For an expense to affect a person’s budget in proportion to its magnitude, an item must be both booked and posted to a meaningful account. Failures to execute either step of the process lead people to account insufficiently for the appropriate size or 1

SUSSMAN, SHARMA, AND ALTER


2

impact of the expense and can cause normative errors, such as overconsumption or underconsumption. Research in mental accounting has shown that people often track their spending by posting purchases to specific categories, and differences in how these categories are formed and evaluated can influence subsequent spending decisions (e.g., Kahneman & Tversky, 1984; Thaler, 1985, 1999; Tversky & Kahneman, 1981). Recent research has demonstrated that people have difficulty placing exceptional expenses into budget categories (Sussman & Alter, 2012). Exceptional (relative to ordinary) expenses are those that are perceived as unusual and infrequent, or irregularly timed. The definition applies to cases where either the occasion or the associated cost is exceptional for a given individual. Exceptional expenses may happen to be rare expenses, but the rarity of the expense alone does not define the item’s exceptionality. Some expenses that are rare in the broader world may in fact seem ordinary if they occur with regularity for a given individual. For example, if someone were to purchase craft beers that are brewed in limited quantities on a weekly basis, these expenses would be considered ordinary for that person. In contrast, people are more likely to construe exceptional expenses as unique occurrences or to consider them as part of isolated or ad hoc budget categories, rather than recognizing them as part of spending on a larger category of goods. For example, people are likely to consider spending money on a weekend with friends visiting from out of town, purchasing a present for a 50th birthday, and buying tickets for a concert tour as unusual expenses that will occur only once or rarely rather than as part of a broader category of expenses (e.g., “spending on infrequent festivities”). When people believe that expenses will either not recur or will recur infrequently, they may fail to book the item in their budget, encoding the infrequent expense as trivial or irrelevant to their budgets at large.

Current Research In the current research, we examined whether people would be more willing to donate to charity when they perceived charitable expenses as relatively more exceptional rather than ordinary. In particular, we examined how exceptional framing might interfere with people’s budgeting processes and in turn alter how much they spend on charitable giving. We propose that reframing donations as exceptional expenses weakens ties to a person’s budget, and, in the context of charitable giving, liberates donors to make larger donations. Although existing research on budgeting for charitable donations is extremely limited, one notable exception is “pennies-aday” framing (Gourville, 1998). This research suggests that framing a single donation as a series of smaller ones (e.g., $1 per day vs. $365 per year) can increase giving by disrupting the budgeting process. The pennies-a-day model works primarily by encouraging people to compare each installment’s small size with other very small expenses rather than the aggregate expense with other large expenses. Similar to pennies-a-day framing, we suggest that exceptional framing disrupts the budgeting process and leads to higher giving; however, we propose that this disruption occurs through a different mechanism. We suggest that exceptional framing disrupts the budgeting process by distinguishing target expenses from more frequent, recurring budget strains. That is, exceptional framing leads people to disregard the cost of expenses

primarily because those expenses seem more distinctive and infrequent. In this way, framing donations as more exceptional (rather than ordinary) may prevent consumers from registering the full magnitude of the expense and in turn influence their willingness to donate. Therefore, we hypothesize that people will respond more generously to charitable appeals when the same donation is framed as more exceptional rather than more ordinary. In addition to our prediction that exceptional framing will enhance charitable behavior, we also suggest that exceptional framing will affect people’s budgeting process. We hypothesize that people will consider the impact of a donation on their budget to a lesser degree when the donation is framed as more exceptional, in part because they are more likely to neglect the consequence of similar donations recurring over time. However, the process required to shift perceptions of exceptionality may be different for encouraging additional donations from existing donors versus encouraging a new donor to give for the first time. For example, someone who gives repeatedly to a charity may have a more concrete understanding of the regularity of a particular charity drive, and it may be harder to shift that person’s perception of whether the donation is ordinary or exceptional. In the examination that follows, we gather data from the general population, most of whom are not current donors to the charities being considered. Thus, we focus our attention on encouraging new donations only, a topic we revisit in the general discussion. We tested our hypotheses in five lab and field experiments. We first tested a manipulation focused exclusively on the frequency of a charitable event in a short advertisement. Participants making donation decisions chose to give more when the charitable appeal was framed as exceptional rather than ordinary (Experiment 1). A field experiment examining Internet users demonstrated that the same ads posted online through Google Adwords yielded a higher click-through rate when the ad was framed as exceptional (Experiment 2). We next examined a potential mechanism underlying the effect of exceptional framing on charitable behavior by identifying differences between how people categorize exceptional and ordinary expenses. We found that people assign exceptional expenses to categories populated by a smaller set of items, as compared with ordinary expenses (Experiment 3). Building on this knowledge, we hypothesized and found that people consider the effect of the donation on their budget to a lesser degree, and thus express a greater likelihood of donating, when the plea is framed in exceptional terms (Experiment 4). We then provided a demonstration of exceptional framing increasing real donations. We showed that people rate their budget as less relevant to their donation decision when the donation is described as exceptional rather than ordinary, and that this budget consideration mediates giving patterns (Experiment 5). We conclude by discussing additional factors contributing to differences in the budgeting process, implications for charitable giving, as well as ways that these findings may generalize to improve welfare in other domains.

Experiment 1 The current experiment focused on altering perceived frequency as a method of varying the exceptionality of an annual event, and tested whether this minimal difference would cause differences in donation decisions.


EXCEPTIONAL FRAMING OF CHARITABLE DONATIONS

3

Method

Results and Discussion

Participants. Participants (N ⫽ 401) were recruited online through the Amazon.com Mechanical Turk platform and completed the experiment for nominal monetary compensation. Participants ranged in age from 18 to 73 (M ⫽ 32), and 33% were female. We chose sample sizes in this study and the remaining studies based on the strength of each manipulation, and the size of the effect we expected it to have on each dependent measure (Simmons, Nelson, & Simonsohn, 2011). Materials. We created two versions (ordinary and exceptional) of advertisements that promoted an Alzheimer’s Association charity walk. Advertisements were modeled after Google Adwords (see Figure 1), which take the form of short messages with a maximum of 25 characters in the headline, 35 characters in two description lines, and a URL. The messages differed only in one line of text designed to manipulate the walk’s perceived exceptionality: “only once a year” (exceptional) or “held annually” (ordinary). We pretested these messages using a separate pool of participants to ensure that the messages differed on the critical exceptionality dimension (i.e., perceived frequency) but did not have unintended effects on other dimensions (including the importance of the charity, impact of the charitable gift, beliefs about the ability to donate in the future, and preferences for uniqueness). Procedure. Participants were randomly assigned to view either the ordinary or the exceptional advertisement on their computer. We instructed participants to imagine that they viewed the ad while browsing the Internet, and to read it carefully. The ad appeared immediately below the instructions and remained on the computer screen for a minimum of 10 s, at which point participants could proceed to the next screen when they chose. On the following page, with the advertisement no longer in view, participants indicated first whether or not they would donate to the Alzheimer’s walk if given the opportunity, and then how much money they would donate to the charity walk if given the opportunity. After completing the central dependent variables, participants responded to a series of demographic questions. The final section included an instructional manipulation check (see Oppenheimer, Meyvis, & Davidenko, 2009) to ensure that participants were reading the instructions carefully and would be able to pick up on the subtle differences across experimental conditions.

Prior to analysis, 18 participants were excluded for having taken this or a related survey previously, and 41 participants were excluded for failing the instructional manipulation check. Note that instructional manipulation check failure rates found across our experiments were consistent with previous research (e.g., up to 46% in Oppenheimer et al., 2009). Patterns of results were consistent when we included these participants. Because the data in this and some of the remaining studies were not normally distributed, we conducted nonparametric rather than parametric tests. In each case, parametric tests returned similar results. Participants reported being significantly more likely to donate in the exceptional condition (46%) than the ordinary condition (35%; Mann–Whitney U ⫽ 2.03, p ⫽ .043, r ⫽ .11). Additionally, the average donation amounts were significantly higher in the exceptional than in the ordinary condition (MORD ⫽ $4.82, SD ⫽ 8.34, 95% CI [3.58, 6.05]; MEXC ⫽ $7.13, SD ⫽ 13.62, 95% CI [5.08, 9.19]; Mann–Whitney U ⫽ 2.02, p ⫽ .044, r ⫽ .11). One strength of the current study is that it examined donation behavior in a controlled laboratory setting. However, when people see online advertisements in their natural context, they are overloaded with extraneous information and have limited (if any) attention to dedicate to the ad. Thus, we designed Experiment 2 to examine whether people respond to subtle differences in advertisements in a more natural, ecologically valid context.

Experiment 2 Because Experiment 1 tested how differences in messaging would influence participants in a laboratory setting, Experiment 2 extended findings to examine whether these effects would persist when people encountered the ads in their daily lives. The current experiment used advertisements placed through the Google Adwords platform to analyze how people would respond to ads in the context of their daily Web searches. Most Google Adwords customers pay on a cost-per-click basis, with the advertisements reaching people on more than 2 million Web sites and hundreds of thousands of apps (Google Annual Report, 2012). Across the United States, Google’s search and advertising tools generated $94 billion in economic activity in 2012 (Google Economic Impact Report, 2012). Thus, testing our hypothesis in this real world context was not only interesting from a theoretical standpoint (to establish external validity) but also from a managerial standpoint, as Google ads are a relevant and viable option for charity advertising.

Method

Figure 1. Ordinary (A) and exceptional (B) advertisements modeled after Google Adwords used in Experiment 1 and images of ordinary (C) and exceptional (D) Google Adwords used in Experiment 2. See the online article for the color version of this figure.

Participants. Advertisements were presented on Web searches based in the Northeast, targeting specific locations where the advertised walk would be taking place, and we tracked the click patterns of those who viewed the target ads. Specifically, we tracked the number of times each Google ad appeared online, and the number of clicks on each ad. The data for all individuals who engaged in a Web search that generated the target ad were included in the experiment; there were no exclusions. People were not aware that their responses were being tracked or that they were


4


participating in an experiment. In total, the advertisement appeared on 141,907 Web pages. Materials and procedure. Two sets of short advertisements (ordinary and exceptional) were created through the Google Adwords platform (see Figure 1). These ads had the same text as in Experiment 1, and contained a Web link that directed clicks to the Alzheimer’s Association’s walk Web site. We created a set of about 20 keyword combinations (e.g., “charity walk”) that determined when the ads would appear in searches. These keywords were identical across ads, and all ads appeared on the side of the search results screen. Additionally, we set a daily budget of $10 per day for the cost of displaying the ads, which had a maximum cost of $0.50 per click and an average cost per click determined by an automatic bidding strategy set through the Google Adwords platform. We posted and rotated the ads daily from July through October 2012, when the walk took place, and we tracked the number of times the ads appeared as well as the number of times people clicked the ads. Our dependent variable was the clickthrough rate: the frequency of clicks divided by the number of times the ad appeared online. Click-through rates measure the first action taken in response to Google ads and serve as a proxy for people’s interest in helping.

Results and Discussion Average click-through rates for Google Adwords are estimated to be in the neighborhood of 1% to 3%, although these numbers vary considerably (Testaverde, 2013). In other words, if an ad were to appear on 100 people’s screens, the expectation would be that between one and three people would click on it. In our study, the ordinary ad appeared 76,416 times, and people clicked on links through this ad 1,270 times (1.66%, 95% CI [1.57%, 1.75%]). The exceptional ad appeared 65,491 times, and people clicked on links through this ad 1,204 times (1.84%, 95% CI [1.74%, 1.94%]). Given that the prevalence of ad clicks is low in this environment, we analyzed the effect of exceptional framing on click-through rates using the risk ratio statistic (e.g., Altman, 1991), which compares the ratio of the probability of the event (clicks) occurring in the treatment group to the probability of the event occurring in the control group. In this case, the calculated risk ratio was 1.11, indicating that people were 1.11 times as likely to click on the Google ad when it was framed as exceptional rather than ordinary (risk ratio ⫽ 1.11, 95% CI [1.02, 1.20], Z ⫽ 2.53, p ⫽ .011). Expressed differently, for every set of 550 people, the exceptional advertisement would receive one more click than the ordinary ad. Given the large volume of views associated with Google AdWords, and the ease of implementing such minor wording changes, practitioners may meaningfully increase donor engagement regardless of the manipulation’s small effect size. These results provided additional evidence that subtle variation in the exceptionality of the charitable activity through manipulation of perceived frequency influences how people respond to charitable pleas. Exceptional framing led to increased interest in the charity walk, measured by the percent of people who clicked on an advertisement when it appeared on their screen. Clickthrough rates measure initial interest in charity appeals, and each additional click provides charities with an opportunity to spread their message and attract new donors, an opportunity that never becomes available if those potential donors choose not to click on

the advertisements. Although this study did not measure donation amount (as was measured in the prior study), the current experiment provides evidence that small changes in descriptions of the charity walk differentially influence behavior in a field setting. Coupled with findings from Experiment 1, the existing data demonstrate both that people express greater interest in and that they will donate more money to a charity in response to an exceptionally framed appeal.

Experiment 3 Thus far we have established that exceptional framing boosts donations, and in Experiment 3, we aimed to examine the process. Prior research on charitable donations suggests that donors may consider a charitable contribution as drawing from a specific mental account for philanthropy. However, mental accounts are known to be flexible, and donors may thus construe donations as drawing from different mental accounts depending on the context (Kahneman & Knetsch, 1992; LaBarge & Stinson, 2014). This malleability of mental accounts suggests that there is room for shifting how (and from which account) people budget for donations. Consistent with previous research on exceptional expenses (Sussman & Alter, 2012), we hypothesized that people considering a charitable donation framed as exceptional would construe that donation as part of a more sparsely populated category of goods. Because the mental accounting literature shows that people set a certain budget for each mental account, rather than a budget per item in those accounts (cf. Cheema & Soman, 2006), they should be willing to spend more on items that belong to sparsely populated accounts. In this way, categorizing the target item within a smaller rather than larger set of items would leave more room in the exceptional item’s budget category to spend on the exceptional item. Placing an item alongside fewer other items is also consistent with beliefs that the item is part of a more exclusive category, potentially contributing to an even higher valuation (e.g., Groth & McDaniel, 1993; Lynn, 1991). Thus, this categorization can lead people to spend more on each item than they would if they considered the same item to be part of a larger group of goods. Additionally, people might be less likely to record items at all, to the extent that they have difficulty placing purchases into a broad budgeting account. Experiment 3 was designed to examine how exceptional framing influences categorization of charitable behaviors. Consistent with previous research on exceptional expenses (Sussman & Alter, 2012), we hypothesized that people considering a charitable donation framed as exceptional will construe that donation as part of a smaller category of goods. Additionally, the current experiment aimed to extend findings from a specific charity event (i.e., the charity walk) to donations solicited without a particular linked occasion. The experiment held donation wording identical and altered exceptionality by varying perceived similarity to other items, examining both how the donation is categorized and maximum willingness to donate as a consequence of the variation.

Method Participants. U.S. residents (N ⫽ 400) were recruited online through the Amazon.com Mechanical Turk platform and completed the experiment for nominal monetary compensation. Par-



5

ticipants ranged in age from 18 to 73 (M ⫽ 30), and 61% were female. Materials and procedure. Participants were told, “We are interested in understanding how people think about expenses. Below are some expenses a person might incur. We’d like you to think about how those expenses might be either similar to or different from other expenses and write your thoughts in the text boxes provided.” They then saw two different items, first an airplane ticket—intended as a distractor—and then a donation to charity. For one item, participants described how the item was “common or similar to other expenses,” and for the other, they described how the item was “unique or different from other expenses.” The pairing of the task and item was randomly determined. In the ordinary condition, the similarity description task was paired with the donation, and in the exceptional condition, the uniqueness description task was paired with the donation. On the following page, participants reviewed a list of eight expenses that they might encounter in their daily lives, presented in a random order. For example, the items included a gift for a good friend, organic milk from a local supermarket, and running sneakers (see Appendix A for a complete list). One item on the list was a donation to charity. Participants were told to sort the items into meaningful, budget-relevant groups. They read the following:

reported the donation to be significantly more similar to other expenses (MORD ⫽ 3.29, SD ⫽ 1.72, 95% CI [3.04, 3.54] vs. MEXC ⫽ 2.38, SD ⫽ 1.39, 95% CI [2.17, 2.58], t(354) ⫽ 5.52, p ⬍ .001, d ⫽ .58, 95% CI [0.37, .80]), whereas those in the exceptional condition reported the donation to be significantly more different (MORD ⫽ 4.78, SD ⫽ 1.70, 95% CI [4.53, 5.03] vs. MEXC ⫽ 5.47, SD ⫽ 1.31, 95% CI [5.28, 5.66], t(354) ⫽ 4.28, p ⬍ .001, d ⫽ .46, 95% CI [0.24, 0.67]). Consistent with our hypothesis, participants in the exceptional condition placed the donations into smaller groups (M ⫽ 1.63, SD ⫽ .78, 95% CI [1.51, 1.75]) than did those in the ordinary condition (M ⫽ 1.82, SD ⫽ .91, 95% CI [1.69, 1.95], t(352) ⫽ 2.13, p ⫽ .034, d ⫽ .22, 95% CI [0.02, 0.43]). Participants were more likely to group the exceptional items with fewer other items when considering budget-relevant categories. After having shown how exceptional framing influences the categorization of items when forming budgets, we next examined how exceptional framing may influence budget considerations directly. We propose that people will consider a donation to be a smaller setback to their overall budget when it is framed as exceptional, and we test this proposition in Experiment 4.

. . . Imagine that this list includes everything you have spent money on over the past month, except for essential spending (rent, groceries, etc.). Your task is to sort these items into groups so that the items in each group seem similar enough that they belong together, and so that they will be useful for budgeting purposes. For example, if you were to see the items: office chair, desk, couch, you might group the first two together in an “office furniture” pile, and separate the third into a “home furniture” pile if you think it would be useful to consider how much you spend in each of these groups. Alternatively, you could put all three together as furniture, or separate each of the three into its own group if you see no overarching connections. There are no correct answers, and you should group the items as you see fit . . .

Taken together, the experiments presented so far demonstrate that exceptional framing increased a variety of charitable behaviors including the likelihood of clicking on the charity’s advertisement during routine Web browsing, and actual donations to the charity. They also suggested that the perceived frequency of exceptional relative to ordinary events is at least partly responsible for these effects and provided initial evidence that exceptional framing alters the budgeting process. Specifically, findings in Experiment 3 showed that thinking of an identical expense as exceptional influences the posting process (cf., Heath, 1995; Heath & Soll, 1996), in which people choose how to categorize the expense within their budget. In the next experiment, we aimed to replicate prior giving patterns and explore further the psychological mechanism. Our objective was to understand whether exceptional framing may also influence the booking process, in which people choose whether to record the expense within their budget initially. We hypothesized that people would be less likely to consider the effect of a donation on their budget when it was framed as exceptional. To test this hypothesis, we presented participants with one of two donation appeals from a small healthrelated charity and measured how likely they were to consider their budget alongside their donation.

After completing the categorization task, participants were told that we were interested in how much they would spend on a subset of the items they categorized, and they stated the highest amount that they would pay for three of the items: a donation to charity and two distractors (organic milk and movie tickets). Finally, participants responded to two manipulation checks. The first asked, “As you completed this study, to what extent did you think that a charitable donation is an expense that is common or similar to other expenses?” with responses on a scale ranging from 1 (not at all similar) to 7 (extremely similar). The second asked, “As you completed this study, how unique or different did you think a charitable donation is to other types of expenses?” with responses on a scale ranging from 1 (not at all unique) to 7 (extremely unique). Participants responded to demographic questions along with an instructional manipulation check before exiting the survey.

Results and Discussion Prior to analysis, 18 participants were excluded for having previously taken this or a related survey, and 26 participants were excluded for failing the instructional manipulation check. Results were qualitatively the same when we included these participants. Manipulation checks confirmed that the independent variable had its intended effect. Participants in the ordinary condition

Experiment 4

Method Participants. U.S. residents (N ⫽ 199) were recruited online through the Amazon.com Mechanical Turk platform and completed the experiment for nominal monetary compensation. Participants ranged in age from 18 to 62 (M ⫽ 29), and 67% were female. Materials and procedure. Participants were randomly assigned to the ordinary or exceptional condition. All participants were told the following: “Imagine that you get a flyer in the mail for Alex’s Lemonade Stand, a charity dedicated to helping fight childhood cancer. Donating to charity is important to you, and you


6


view this as a worthy cause, so you read the flyer closely.” Then, in the ordinary condition, participants read (in standard font): “This mailing is part of a regular charity drive that happens annually. The charity is requesting a donation every year going forward.” While those in the exceptional condition read: “This mailing is part of a special charity drive that happens only once a year. Alex’s Lemonade Stand is requesting only one donation a year going forward.” The text was presented alongside a corresponding banner ad (see Figure 2). Participants were then asked, “How likely would you be to agree to this donation request today?” and responded on a scale from 1 (not at all likely) to 7 (extremely likely). On the following page, participants were asked: “Given that the donation occurs [every single year/only once a year], to what extent would you consider the effect of this donation on your budget?” and responded on a scale ranging from 1 (not at all) to 7 (very much). Differences between conditions were highlighted in the text only; participants saw all text in standard font. Next, to ensure that the manipulation had its intended effect of varying perceived frequency, we asked participants the following: “Based on the passage above, how often do you think the charity drive associated with the mailing occurs?”, and responded on a scale ranging from 1 (very infrequently) to 7 (very frequently). Then, to confirm that the manipulation did not alter perceptions on other potentially relevant dimensions such as perceived importance or scarcity, we asked participants, “How important do you think that the mission of Alex’s Lemonade Stand is?” and “How rare of a charity do you think that Alex’s Lemonade Stand is?” and responded to each on a scale from 1 (not at all) to 7 (extremely). Participants responded to demographic questions along with an instructional manipulation check before exiting the survey.

Results and Discussion Prior to analysis, seven participants were excluded for having previously taken this or a related survey, and 13 participants were excluded for failing the instructional manipulation check. Patterns of results were consistent when we included these participants. The manipulation checks confirmed that participants viewed the charitable appeal as less frequent in the exceptional condition (MEXC ⫽ 2.60, SD ⫽ 1.65, 95% CI [2.26, 2.94] vs. MORD ⫽ 3.88,

SD ⫽ 1.74, 95% CI [3.52, 4.24], t(177) ⫽ 5.02, p ⬍ .001, d ⫽ .75, 95% CI [0.46, 1.06]), but that they did not view the charity as more important (MEXC ⫽ 5.42, SD ⫽ 1.42, 95% CI [5.13, 5.71] vs. MORD ⫽ 5.57, SD ⫽ 1.29, 95% CI [5.30, 5.84], t(177) ⫽ .74, ns) or rare (MEXC ⫽ 2.86, SD ⫽ 1.50, 95% CI [2.55, 3.17] vs. MORD ⫽ 2.72, SD ⫽ 1.46, 95% CI [2.41, 3.03], t(177) ⫽ .64, ns) across conditions. Consistent with our hypothesis, participants in the exceptional condition reported being significantly more likely to donate to the charity than did those in the ordinary condition (MEXC ⫽ 4.19, SD ⫽ 1.81, 95% CI [3.82, 4.56] vs. MORD ⫽ 3.43, SD ⫽ 1.77, 95% CI [3.06, 3.80], t(177) ⫽ 2.82, p ⫽ .005, d ⫽ .42, 95% CI [0.12, 0.72]). Additionally, participants in the exceptional condition considered the effect of the donation on their budget to a lesser degree (MEXC ⫽ 3.48, SD ⫽ 1.78, 95% CI [3.11, 3.85] vs. MORD ⫽ 4.56, SD ⫽ 1.98, 95% CI [4.15, 4.97], t(177) ⫽ 3.82, p ⬍ .001, d ⫽ .57, 95% CI [0.27, 0.87]). This experiment replicates and extends findings in Experiments 1, 2, and 3 by showing that exceptional framing enhances charitable behavior, in this case measured by donation likelihood. Additionally, it provides the first evidence that framing an identical contribution as exceptional directly influences the amount that people consider their budget.

Experiment 5 Earlier experiments showed that framing donations as exceptional influenced behavior as measured through stated willingness to donate (Experiment 1) and through clicks on advertisements framed as exceptional in a real online environment (Experiment 2). Additionally, they showed that exceptional framing changes the way people categorize the donation in their budget (Experiment 3) and reduces the likelihood that people will consider the effect of the donation on their budget (Experiment 4). Experiment 5 was designed to strengthen and extend findings from our previous studies in a few ways. First, we used stimuli similar to that used in Experiment 4, but we explicitly informed participants in both conditions that the charity flyer is the only one sent by the charity, thus implying that participants should not expect to receive additional solicitations from the charity that year. Second, we relied on an incentive compatible design with real behavior in which participants’ donation decisions were implemented. Finally, we aimed to gather process evidence with greater precision. Specifically, we asked people to allocate a constant sum to indicate their consideration of a variety of factors, including the extent to which budgetary factors restrict them from considering the donation. This measure allowed us to identify whether people disregard their budget, even the donation’s affordability is highlighted to them prior to their donation decision.

Method

Figure 2. Ordinary (A) and exceptional (B) banner advertisements used in Experiment 4 and 5. See the online article for the color version of this figure.

Participants. U.S. residents (N ⫽ 1,045) were recruited online through the Amazon.com Mechanical Turk platform and completed the experiment for nominal monetary compensation. Participants ranged in age from 18 to 74 (M ⫽ 33), and 41% were female. Materials and procedure. Participants were randomly assigned to the ordinary or exceptional condition. They were introduced to the charity opportunity by reading the following: “Imag-



ine that you get a flyer in the mail for Alex’s Lemonade Stand, a charity dedicated to helping fight childhood cancer. It is the only flyer sent by the charity, and the charity sends it [every year/once a year]. Since donating to charity is important to you, and you view this as a worthy cause, you read the flyer closely. It happens to be part of a charity drive that happens [annually/only once a year]. You are considering donating to the charity drive today.” Participants then saw a banner advertisement that described the charity event as being [held every year/only once a year] (see Figure 2). Participants in the ordinary condition saw the first phrase in each of the brackets above while those in the exceptional condition saw the second phrase. Next, participants allocated 100 points to different factors that might influence their donation decisions based on the relative importance of each factor in shaping these decisions. The target budgeting-related item that served as our process measure was the perceived affordability of the donation (included last on the list). The additional filler items included: the importance of the charitable cause, the general importance of giving to charity, and a more general “other” category. On a separate page, participants were told that three people would be randomly selected to win an Amazon gift voucher worth up to $50. They were given the opportunity to donate all or a portion of that voucher to Alex’s Lemonade Stand. We counterbalanced the order in which the process measure (donation affordability) and dependent measure (donation amount) were administered. To ensure that the manipulation had its intended effect of reducing perceived frequency, participants read: “Based on the flyer you viewed earlier, how often does it feel as though the charity drive associated with the mailing occurs?” and “Based on the flyer you viewed earlier, how often does it feel as though the Lemon Ball occurs?”, and responded on a scale ranging from 1 (very infrequently) to 7 (very frequently). They also answered the following question: “Based on the flyer you viewed earlier, how unusual do you consider the Lemon Ball for a charity event?” and responded on a scale ranging from 1 (not at all unusual) to 7 (extremely unusual). Then, to confirm that the manipulation did not alter perceptions on other potentially relevant dimensions such as perceived importance or scarcity, participants were asked, “How important do you think that the mission of Alex’s Lemonade Stand is?” and “How rare of a charity do you think that Alex’s Lemonade Stand is?” and responded to each on a scale ranging from 1 (not at all) to 7 (extremely). Participants responded to demographic questions along with an instructional manipulation check before exiting the survey.

Results and Discussion Prior to analysis, 65 participants were excluded for having previously taken this or a related survey, and 18 participants were excluded for failing the instructional manipulation check. Patterns of results were consistent when we included these participants. Manipulation checks confirmed that participants viewed the charity drive and the charity event as less frequent (averaged response, MEXC ⫽ 2.05, SD ⫽ 1.29, 95% CI [1.93, 2.17] vs. MORD ⫽ 3.11, SD ⫽ 1.55, 95% CI [2.97, 3.25], t(960) ⫽ 11.48, p ⬍ .001, d ⫽ .74, 95% CI [0.61, 0.87]), and more unusual (MEXC ⫽ 3.57, SD ⫽ 1.73, 95% CI [3.42, 3.72] vs. MORD ⫽ 3.12, SD ⫽ 1.65, 95% CI [2.97, 3.27], t(960) ⫽ 4.13, p ⬍ .001, d ⫽

7

0.27, 95% CI [0.14, 0.39]) in the exceptional condition relative to the ordinary condition. They also confirmed that there were no differences in the perceived importance or rareness of the charitable cause (ts ⬍ 1, ps ⬎ .60). Consistent with the findings from Experiments 1– 4, participants in Experiment 5 donated more of their lottery winnings to the charity in the exceptional condition than in the ordinary condition (MEXC ⫽ 12.93, SD ⫽ 10.93, 95% CI [11.23, 13.19] vs. MORD ⫽ 10.26, SD ⫽ 9.60, 95% CI [9.40, 11.12]; Mann–Whitney U, Z ⫽ 2.59, p ⫽ .010, r ⫽ .20). Next, we examined participants’ responses regarding the donation’s affordability. As expected, participants in the exceptional condition placed lower importance on this measure than did participants in the ordinary condition, indicating that affordability was less of a concern for those in the exceptional condition (MEXC ⫽ 7.36, SD ⫽ 22.95, 95% CI [5.30, 9.42] vs. MORD ⫽ 10.64, SD ⫽ 26.78, 95% CI [8.25, 13.03]; Mann–Whitney U, Z ⫽ 2.62, p ⫽ .009, r ⫽ .21). These results suggest that, even after the question itself encouraged participants to consider the affordability of a donation explicitly, those in the exceptional condition were more likely to discount the possibility that donating would influence their budget at all. To further investigate whether and how budget considerations influenced the donation decision, we examined whether budget considerations mediated the observed pattern of results using a nonparametric, bootstrap resampling approach (Hayes, 2013; Preacher & Hayes, 2008). Specifically, we used Hayes’ (2013) PROCESS macro with bias-corrected confidence intervals based on 20,000 bootstrapped samples. We found support for the suggested mediation, as the 95% confidence interval for the indirect effect of condition (ordinary ⫽ 0, exceptional ⫽ 1) on dollars donated via the budget consideration mediator did not include zero (0.001, 0.017). These results suggest that exceptional framing decreases the consideration of one’s budget through overall affordability perceptions, which in turn increases maximum willingness to donate. In other words, when the donation is framed as exceptional, people are less likely to believe that budget considerations limit their ability to engage with the donation decision, causing them to donate higher amounts on average.

General Discussion Five lab and field studies provided converging evidence that exceptional framing enhances charitable behavior and lessens people’s consideration of the donation on their budget. These effects were observable with subtle manipulations that used a minimal difference in framing focusing exclusively on the frequency of the charitable event. Notably, this is the first work to systematically parse the psychological mechanism driving exceptional framing. We demonstrate that exceptional framing decreases the relevance of budget considerations and the likelihood of considering an item’s affordability. This work is also the first to demonstrate practical applications of exceptional framing and test those applications in a field setting. The current research extends understanding of charitable donations in a variety of ways. First, it demonstrates that altering exceptionality directly influences the extent to which people mentally account for charitable giving decisions. Second, it shows that the same item or event, occurring with the same objective fre-


8


quency, can be described as relatively more exceptional, thereby increasing charitable giving. Additionally, it provides evidence for this pattern for consequential donation decisions as well as in a large field experiment. Thus, another key contribution of this work is a strategy that charities can easily implement to alter the perceived exceptionality of the charitable plea. From a theoretical perspective, the current work provides a glimpse into the mental accounting processes associated with giving decisions, a topic about which relatively little is known. Although previous research has examined egoistic, altruistic, and emotional drivers of giving, existing literature is largely silent on budgeting factors, with literature on pennies-a-day being one exception. We present one of the first demonstrations that altering exceptionality directly influences the extent to which people consider their budgets when deciding whether and how much to give. Exceptional framing disrupts the budgeting process by making people consider a donation, cause, or event as infrequent and less consequential for their budget. People are more likely to categorize those exceptionally framed donations in a mental account with fewer other items, and less likely to dismiss the idea of donating as a result of budget constraints.

Conclusions and Future Directions The current research uses insights from prior work on charitable giving and exceptional expenses, and builds and extends understanding in a variety of ways. Previous charity research has identified ways in which factors related to the donor (e.g., Sharma & Morwitz, 2015; Kappes, Sharma, & Oettingen, 2013; Small, 2011), the recipient, and the charitable context (Bendapudi, Singh, & Bendapudi, 1996) contribute to charitable giving. However, existing research has devoted less attention to how people mentally account for charitable gifts. In addition, the previous work on exceptional expenses has focused mostly on how a bias in framing exceptional expenses can harm individuals. The current research showed that this bias can also reap societal benefits by encouraging donations. Charities have a constant need to raise funds for their causes on an ongoing basis. One common way to address this need is head-on through a consistent stream of messages that emphasize the repetition and regularity of charitable events and drives. For example, charities often use strategies that explicitly highlight the stability and tradition of both the charity and the donation plea (e.g., annual walks, galas, or fundraising campaigns). While it might be important to have charitable fundraisers held on a regular basis, highlighting their regularity (e.g., through the “annual” tag) may emphasize that the donations are both more ordinary and more financially depleting. Indeed, the work presented here suggests that presenting charitable pleas as more infrequent and distinct rather than frequent can be more effective. Manipulations used in the context of this article held the amount of exposure to charitable solicitations and fundraising events constant, varying only perceived exceptionalness, which included differences in subjective frequency. This approach suggests a concrete method for practitioners aiming to increase donations from new donors that can be operationalized through a minimal difference in the wording of the charitable appeal. It has the potential to be effective in encouraging donations without the risk of limiting donations through reduction of fundraising attempts (e.g., sending solicitations less frequently). However, it seems

plausible that manipulating the actual frequency of charity mailings or even charity events could be even more effective in altering perceptions of exceptionality and thus increasing donations. The data here leave several new questions for investigation. In particular, it will be worthwhile to understand how the patterns observed here unfold for existing donors over time. Irrespective of ongoing donations, charities that offer lifetime rewards—such as membership to a museum—in exchange for large, one-time donations, could potentially benefit from emphasizing the one-time nature of the expense. Understanding how charitable giving unfolds over time is a complex issue. Research will be needed to learn whether increased giving in response to one solicitation results in lower donation likelihood in response to the next charitable plea, or whether people adapt to exceptional framing given increased exposure to it. Future research should also explore additional budgeting factors that may contribute to this pattern. The notion that people are more likely to neglect opportunity costs (cf., Frederick, Novemsky, Wang, Dhar, & Nowlis, 2009), an indirect measure of budget considerations, when encountering an exceptional donation plea, could lead to the observed patterns. Exceptional framing may also influence the budgeting process in other ways, for example by increasing the importance of the donation’s budget category. Additionally, for the sake of consistency across experimental conditions, we systematically varied the manner in which information was presented while holding the information conveyed constant. Therefore, the majority of cases examined in the article included mention of an event’s frequency (e.g., “only once a year”). However, variations that exclude any mention of frequency—and hence do not draw attention to its recurrence— may be even more effective in encouraging donors to consider a charitable gift or behavior as a one-time exception, which could, in-turn, lead to an even greater increase charitable behavior. Finally, although the current work focuses on charitable giving, the budgeting mechanism suggests the findings may have broader applications to a range of purchases and can be leveraged to encourage beneficial behaviors in a variety of ways. For example, supermarkets could describe healthy foods such as fruits and vegetables as more exceptional (e.g., by advertising a particular variety as being available “only in the summer” vs. “every summer”) to encourage healthier eating. Or, banks could emphasize events already considered exceptional as opportunities for making affordable, one-time savings deposits for the self (e.g., tax refunds, bonuses) or others (e.g., child’s birthday, graduation) to increase savings. Extensions to these and related areas would be worthy of future investigation. Charities rely heavily on individual donors to operate and help progress societal goals, yet little is known about how the framing of charitable appeals affects people’s mental budgeting processes and hence their willingness to give. The current work addresses this gap by demonstrating that exceptional framing enhances charitable behavior via the reduction of budget considerations. This work provides both charities and researchers with novel insight into the psychology of consumer budgeting and giving. In addition, given that charitable behavior tends to increase happiness for donors (e.g., Dunn & Norton, 2013),


using exceptional framing to enhance charitable behavior has the potential to benefit both the donor and the recipient.


References Altman, D. G. (1991). Practical statistics for medical research. London: Chapman and Hall. Andreoni, J. (1989). Giving with impure altruism: Applications to charity and Ricardian equivalence. Journal of Political Economy, 97, 447–1458. http://dx.doi.org/10.1086/261662 Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow giving. The Economic Journal, 100, 464 – 477. http://dx.doi.org/10.2307/2234133 Batson, C. D. (1990). How social an animal? The human capacity for caring. American Psychologist, 45, 336 –346. http://dx.doi.org/10.1037/ 0003-066X.45.3.336 Bendapudi, N., Singh, S. N., & Bendapudi, V. (1996). Enhancing helping behavior: An integrative framework for promotion planning. Journal of Marketing, 60, 33– 49. http://dx.doi.org/10.2307/1251840 Burnett, J. J., & Wood, V. R. (1988). A proposed model of the donation decision process. Research in Consumer Behavior, 3, 1– 47. Cheema, A., & Soman, D. (2006). Malleable Mental Accounting: The Effect of Flexibility on the Justification of Attractive Spending and Consumption Decisions. Journal of Consumer Psychology, 16, 33– 44. http://dx.doi.org/10.1207/s15327663jcp1601_6 Dickert, S., Sagara, N., & Slovic, P. (2011). Affective motivations to help others: A two-stage model of donation decisions. In D. M. Oppenheimer & C. Y. Olivola (Eds.), The science of giving: Experimental approaches to the study of charity. New York, NY: Psychology Press. http://dx.doi .org/10.1002/bdm.697 Dunn, E. W., Aknin, L. B., & Norton, M. I. (2008). Spending money on others promotes happiness. Science, 319, 1687–1688. http://dx.doi.org/ 10.1126/science.1150952 Dunn, E., & Norton, M. (2013). Happy money: The science of smarter spending. New York, NY: Simon & Schuster. Frederick, S., Novemsky, N., Wang, J., Dhar, R., & Nowlis, S. (2009). Opportunity cost neglect. The Journal of Consumer Research, 36, 553– 561. http://dx.doi.org/10.1086/599764 Giving USA. Foundation. (2013). Giving USA 2013 Highlights. Retrieved from http://www.givingusareports.org/ Glazer, A., & Konrad, K. K. (1996). A signaling explanation for charity. The American Economic Review, 86, 1019 –1028. Google Annual Report. (2012). Retrieved from http://www.sec.gov/ Archives/edgar/data/1288776/000119312513028362/d452134d10k. htm Google Economic Impact Report. (2012). Retrieved from http://static .googleusercontent.com/economicimpact/reports/EI_Report_2012.pdf Gourville, J. T. (1998). Pennies-a-day: The effect of temporal reframing on transaction evaluation. The Journal of Consumer Research, 24, 395– 408. http://dx.doi.org/10.1086/209517 Griskevicius, V., Tybur, J. M., Sundie, J. M., Cialdini, R. B., Miller, G. F., & Kenrick, D. T. (2007). Blatant benevolence and conspicuous consumption: When romantic motives elicit strategic costly signals. Journal of Personality and Social Psychology, 93, 85–102. http://dx.doi.org/ 10.1037/0022-3514.93.1.85 Groth, J. C., & McDaniel, S. W. (1993). The exclusive value principle: The basis for prestige pricing. Journal of Consumer Marketing, 10, 10 –16. http://dx.doi.org/10.1108/07363769310026539 Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press. Heath, C. (1995). Escalation and de-escalation of commitment in response to sunk costs: The role of budgeting in mental accounting. Organizational Behavior and Human Decision Processes, 62, 38 –54. http://dx .doi.org/10.1006/obhd.1995.1029

9

Heath, C., & Soll, J. B. (1996). Mental budgeting and consumer decisions. The Journal of Consumer Research, 23, 40 –52. http://dx.doi.org/ 10.1086/209465 Kahneman, D., & Knetsch, J. L. (1992). Valuing public goods: The purchase of moral satisfaction. Journal of Environmental Economics and Management, 22, 57–70. http://dx.doi.org/10.1016/0095-0696(92)90019-S Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39, 341–350. http://dx.doi.org/10.1037/0003-066X .39.4.341 Kappes, H. B., Sharma, E., & Oettingen, G. (2013). Positive fantasies dampen charitable giving when many resources are demanded. Journal of Consumer Psychology, 23, 128 –135. http://dx.doi.org/10.1016/j.jcps .2012.02.001 LaBarge, M. C., & Stinson, J. L. (2014). The role of mental budgeting in philanthropic decision-making. Nonprofit and Voluntary Sector Quarterly, 43, 993–1013. 0899764013489776. Linn, A. (2013, June 22). Charitable giving continues to be a victim of recession. Consumer News and Business Channel. Retrieved from http:// www.cnbc.com/id/100829583 Lynn, M. (1991). Scarcity effects on value: A quantitative review of the commodity theory literature. Psychology and Marketing, 8, 43–57. http://dx.doi.org/10.1002/mar.4220080105 Margolis, H. (1984). Selfishness, altruism, and rationality. Chicago, IL: University of Chicago Press. Oppenheimer, D. M., Meyvis, T., & Davidenko, N. (2009). Instructional manipulation checks: Detecting satisficing to increase statistical power. Journal of Experimental Social Psychology, 45, 867– 872. http://dx.doi .org/10.1016/j.jesp.2009.03.009 Oppenheimer, D. M., & Olivola, C. Y. (Eds.). (2011). The science of giving: Experimental approaches to the study of charity. New York, NY: Psychology Press. Preacher, K. J. (2001). Calculation for the chi-square test: An interactive calculation tool for chi-square tests of goodness of fit and independence [Computer software]. Retrieved from http://quantpsy.org Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879 – 891. http://dx.doi .org/10.3758/BRM.40.3.879 Sergeant, E. S. G. (2014). Epitools epidemiological calculators. Retrieved from http://epitools.ausvet.com.au Sharma, E., & Morwitz, V. M. (2015). Saving the masses: The impact of perceived efficacy on charitable giving to single vs. multiple beneficiaries. Unpublished manuscript, Dartmouth College. Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22, 1359 –1366. http://dx.doi.org/10.1177/0956797611417632 Small, D. (2011). Sympathy biases and sympathetic appeals: Reducing social distance to boost charitable contributions. In D. M. Oppenheimer & C. Y. Olivola (Eds.), The science of giving: Experimental approaches to the study of charity. New York, NY: Psychology Press. Soman, D. (2001). Effects of payment mechanism on spending behavior: The role of rehearsal and immediacy of payments. The Journal of Consumer Research, 27, 460 – 474. http://dx.doi.org/10.1086/319621 Soman, D., & Lam, V. W. (2002). The effects of prior spending on future spending decisions: The role of acquisition liabilities and payments. Marketing Letters, 13, 359 –372. http://dx.doi.org/10.1023/A: 1020374617547 Strahilevitz, M. A. (2011). A model of the value of giving to others compared to the value of having more for oneself: Implications for fundraisers seeking to maximize donor satisfaction. In D. M. Oppenheimer & C. Y. Olivola (Eds.), The science of giving: Experimental approaches to the study of charity (pp. 15–34). New York, NY: Psychology Press.


10


Sussman, A. B., & Alter, A. L. (2012). The exception is the rule: Underestimating and overspending on exceptional expenses. The Journal of Consumer Research, 39, 800 – 814. http://dx.doi.org/10.1086/665833 Testaverde, T. (2013, January 22). When is 2% not a good CTR? The relationship of click-through rate & ad position. Retrieved from http://www.wordstream .com/blog/ws/2013/02/22/click-through-rate-by-ad-position Thaler, R. H. (1985). Mental accounting and consumer choice. Marketing Science, 27, 15–25. http://dx.doi.org/10.1287/mksc.1070.0330 Thaler, R. H. (1999). Mental accounting matters. Journal of Behavioral Decision Making, 12, 183–206. http://dx.doi.org/10.1002/(SICI)10990771(199909)12:3⬍183::AID-BDM318⬎3.0.CO;2-F

Triandis, H. C., & Gelfland, M. J. (1998). Converging measurement of horizontal and vertical individualism and collectivism. Journal of Personality and Social Psychology, 74, 118 –128. http://dx.doi.org/10.1037/ 0022-3514.74.1.118 Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211, 453– 458. http://dx.doi.org/10.1126/ science.7455683 Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. http://dx.doi.org/10.1037/0022-3514.54.6.1063

Appendix Complete List of Items Used in Experiment 3 1. A donation to a charity 2. A gift for a good friend 3. Dinner out with friends 4. Organic milk from your local supermarket 5. Running sneakers 6. Vegetables at a farmer’s market

7. Movie tickets 8. A new sports watch Received July 1, 2014 Revision received March 7, 2015 Accepted March 9, 2015 䡲